: VALUES AND VALUE TRANSFERS: A COMMENT ON ITOH Anwar Shaikh Itoh's essay provides a valuable retrospective on recent debates concerning the Marxian theory of value. He points out that the economic crisis enveloping both cap­ itaJist and socialist worlds from the 1970s onward challenged economists to rethink and refine their basic approaches. Among other things. this sparked a resurgence of interest in fundamental issues such as Ihe theory of value. Thus, what appears as an abstraCt discussion is actually rooted in historical events, with political and economic implications for the current period (above. 53-54), Itoh also emphasizes the importance of Sraffa's work in revitalizing the Marxist discussion ofva1ue theory. Sraffa's pathbreak.ing work provided a new platform for a critique of neoclassical economics, even as it "rchabilitatcd the objcctive theory of value" which is characteristic of me classical and Marxian traditions. But his narrow focus and elliptical style left open the question of the relation between his approach and that of more general economic framewocks. Not swprisingiy, many of his followers set out to incol']X)rate his unconventional approach into a more orthodox "neo-Ricardian" framework, in which the emphasis was on mathematical and functional analyses of CQuilibriwn prices. Inevitably, this gave rise to a divergence between nco-Ricardian and Marxian writers (54--60). Two issues are highlighted by Itoh: the problems which arise with Marx's procedure for linking labor value and money price magnitudes (the transfonnatiOil problem), problems which appear LO undermine the Marxian claim that value magnitudes arc the foundation of price magnitudes; and the claim that value categories are in any case redundant, since market prices actually gravitate around prices of production, not labor values (60). In his survey of the debates around these two axes, ltoh makes many insightful comments and criticisms. He also presents his own solutions to certain key problems raised in the debates, such as those involved in joint production and heterogeneous labor. Here, though, I focus on the most basic issue: the transformation problem and its implications. ~ .1.- r--: vALOE CONTROVERSY 77 One difficulty with Itoh's presentation is that his survey focuses principally on works written before 1980.1 This is problematic for two reasons: first, the lil.crature on the subject burgeoned in the 19705 and 1980s; second, his pre·198O focus presents some difficulties for my own commentary. since most of UJe arguments upon which I rely were advanced afl.cr that date, and are therefore not addressed by Itoh? In what follows, I will begin with points of agrcemenl, move on to the general nature ofIlOh's propJscd solution to the transformation problem, and then contrast it with my own alternative approach.' One crucial pointofagrcementis that the IfanSformation problem is an analytical issue. not a historical onc. The real process of capitalist competition oJ:Crates on 3Ctu.ll market prices, not 00 prices proportional to labor values nor even on prices of production. A market price embodying a higher than average profit rate attracts a correspondingly more rapid influx of capital into the industry, until supply expands more rapidly than demand, pushing d()'A.llUJe market price and hence the profit rate, As this occurs, the influx of capiul begins to slow down Wltil eventual ly supply grows mOte slowly than demand, and lhe price and hence the profit rate ocgin to rise. The overall process is one of pcrpcutal oscillation. in which !he industry's a\o'ernge rate o f profit fl uctualCS around the social average, in what Marx calls "an anai-. chical movement in which rise is compensated by fall and fall by rise" (Marx 1972, 174-75; Shaikh 1984,48-49). There is never any stale of equilibrium in which market pn'ces "converge" to prices of production, Thus prices of production never exist as such. A second important point of agreement is that labor values are dimensionally distinct from prices o f any son, since the fOl1l1er ace in hours of socially ncccssary labor time, while the latter 3re we foons of value expressed in moocy terms. Of course, this merely implies that in order to compare the "sum of values" (hours) with the "sum of prices" (S), or "total surplus value" with "total profit," one must conven onesct into the units of the other (Shaikh 1977, sees. IV- V; 1984.43-44; IlOh 1980, 70-71; above, 61-62). This JX) int is at the heart of our alternative approaches to the the transformation problem. [will return to it shortly. Prices of production function as we inner regulalOrs of market prices. It is Marx's contention that labor values in lum regulate prk:cs of production. just as surplus value regulates nomlal profit.4 To demonstrate th is, Marx proposes to show that one can analytically derive prices of production by beginning from a siLuation in which prices are proportional to labor values and profilS proportional 1O surplus value, and then redistributing profits among industries so as to achieve equal rates of profit. But Ntarx's procedure involves a well·known incompleteness, in that the prices of production he derives are themselves based on input costs still expressed in (prices proportional to) labor values. This deficiency was eventually remedied by showing that Marx's procedure may be thought of as the first stcp in an iterative process of calculation whose end result is exactly !.he "fully tr.msformcd" prices of production of Bortkiewicz or Sraffa (Itoh above. 60). Thus, in the end, prices or : 78 RADICAL ECOKO}.1]CS production (and any other prices as well) arc simply the expression of some redistribution of values and surplus values. Completing Marx's transformation procedure demonstrates the link betwccn his initial derivation of prices of production and the subseqUCI'lt BortkicWK:z "simullancous solution" method (Sweezy 1964, ch. V). But lhis does not directly address an apparently more intractable problem: in any full tmnsfonnation procedure, one cannot maintain both the sum of prices and the sum of profits at their initia1lcvcls (i.e., at levels proportional to the sum of values and the sum of surplus values, respectively). Note th:n lhe transfonnation problem involves the comparison of L..... ,o dj(Ccrent sets of relative prices for a gi'.en set of products produced with a given set of material inputs, labor hours and consumption bundles for workers. Thus the surplus product is also given. As long as !he system is in a self-reproducing Slate, i.e .• in simple or expanded reproduction. the sum of prices represents the purchase price of the total product. Thus keeping the sum of prices constant when considering different seLS of relative prices is tantamount 10 keeping the general purchasing power of money constant. But different seLS of relative prices will yield different measures of (constant 5) money profit, even though the surplus prextuct and hence the total amount of surplus value is WlChanged It therefore appears as if the amount of (constant $) profit is in general independent of both lhc surplus product and surplus value. This is the centra] problem raised by !he tr.lnsfonnation qucsuon. The same result obtains if we keep the general value of money constant instead. A given sum of money prices purchases the lOlaI output, in which the sum of labor values is materialized. 1lle general pwdlaSing power of money over labor value is defmed by the mUo of the sum of the money prices of lOlaI Output (5) to the corresponding sum . of labor values (hours). The reciprocal of this is what Marx dcfmes as the value of money. Multiplying any sum of money by the value of money allows us to translate it into the generallabor-value-commanded by this quantity of money. In this sense, the "sum of (translated) prices" of the whole product will always equal the sum of values of that product. regardless of what happens across an y transformation procedwe(i.e., whether or nO[ the sum of moncy prices or money profiLS or any other sum is kept constant). But here too, as in the earlier case of a constant use-value purchasing power of money, the "sum of (translated) profiLS" will vary independently of surplus value, as relative prices change. Tables lA-2A in Appendix 1 illustrate the basic problem. Table lA is identical to Itoh 's Table 1 of prodUCed labor values in a schema of simple reproduction ([mh 1980, 75), while Table 2A is his Table 2 of money price of production translated inw generallabor-value-commandcd by the value of money m.' Although both tables ha ..'e the same sum of prices, the sum of profits in Table 2 (175 hours) differs from I.hc sum of surplus values (200 hours) in Table I. This is solely the effect of different sets of relative prices applied to the same set of inpuLS and outputs. The question is: what is the impon of this result? On the surface, the above result seems to suggest that circulation can create or THE VALUE CONTROVERSY 79 desaoy value, which conU:ldicts the M:lrxist claim thai value and surplus vaJue are created only in production. For lhis reason, many Marxists have wrestled wilh !his problem, coming up with.3. variety of answers (some of which are surveyed by Itch). llOh himself claims 10 have a new solution, which he lays out in his books (1980, 74-76; 1988,220-26) and swnmarizcs in his sun·cy above. In cffect. ltoh switches the focus of altcntion away from lhc discrepancy betwCCll produced values and realized values (i.e., between Tables IA-18) I() flu! itkmiry of aggrtgmt physicaijlows which is implicir in any s~lf·r~producing S/a/e.